A scratch on a carbon bike frame: when it is cosmetic and when it is not

A scratch on a carbon bike frame is not a single category. Under examination it splits into two distinct cases, and the split decides whether the mark is logged as cosmetic in the inspection report or whether it warrants subsurface NDT before riding again.

The first case is paint-and-clearcoat damage on a non-junction surface, with no surrounding paint bubbling or localized depression. The report logs this as Serviceable in the three-tier safety categorization (VéloColour): no structural damage, but minor cosmetic chips, superficial scratches, or non-structural wear are logged for future monitoring.

The second case is damage that reaches the laminate, sits in one of the four high-stress junctions, or shows raking-light bubbling or rippling around the scratch. This case warrants subsurface NDT before riding, because internal delamination can sit entirely beneath an undisturbed clearcoat. A clean paint surface guarantees nothing; that corollary is exactly what the subsurface methods exist for.

The methods that separate the two cases are stacked in a specific order, from cheapest to most definitive.

The triage protocol for a scratch#

The first method is raking-light examination. The frame is wiped clean to remove dirt and film that can mask fine surface disruptions, then swept under a high-intensity, shallow-angle light source (Carbon Bike Doctor). The shallow angle is what makes the method work: light skimming the surface at a low angle casts micro-shadows from surface relief that flat overhead lighting flattens out entirely.

What raking light reads around a scratch:

• Surrounding paint bubbling, which can indicate that the impact producing the scratch also produced subsurface change.
• Clearcoat rippling adjacent to the scratch, with the same implication.
• Localized depression near the scratch site, which suggests a denting impact rather than a glancing surface contact.
• Surface continuity across the scratch line, where the question is whether the clearcoat is cracked along the line or just abraded across it.

A scratch with no surrounding bubbling, no ripple, no localized depression, and an abraded-rather-than-cracked clearcoat reads as low risk on raking-light triage alone. A scratch with any of those four features moves to the next step.

The second method is digital microscope or optical loupe examination, which separates a benign clearcoat scratch from an active laminate fracture. A clearcoat scratch looks like a discrete cut through a thin uniform layer, with the underlying carbon weave intact and visible beneath the cut. An active laminate fracture looks like a broken weave with disturbed fiber ends, altered light scatter, and often a wider visual footprint than the surface mark suggested. The two are visually distinct under magnification in a way they are not to the unaided eye.

High-stress junctions get this microscope examination by default, not by escalation. The four critical zones are the bottom bracket cluster, head tube transitions, seatpost insertion area, and dropouts. A scratch in one of these zones gets the microscope step regardless of what raking light showed; the redundancy in the ply structure at the junctions can mask damage that becomes visible only under magnification.

When the scratch warrants subsurface NDT#

The triggers for subsurface NDT are specific. Surface triage decides whether the methods below are appropriate; the methods below answer the structural question.

Fluorescent dye penetrant testing is the method for surface-breaking micro-cracks where direct ultrasonic contact is impractical (Target Composites). A low-viscosity dye seeps into the crack, the surface is cleaned, and under UV light the dye trapped in the crack fluoresces, mapping the fracture's exact size and direction. This is the method built for the hairline-crack case: the scratch on the seatpost insertion zone that the digital microscope showed might be a fracture, or the mark on a complex junction where ultrasound can't reach.

Ultrasonic testing is the workhorse for subsurface scanning on conventional tube geometry. A portable transducer coupled to the frame sends high-frequency sound into the laminate, and the system measures transit time to back-wall reflection to resolve wall thickness down to roughly one one-thousandth of an inch (Evident / Ruckus Composites). When the wave hits a void, resin pocket, or delaminated layer under the scratch zone, it reflects early and produces an abnormal echo on the technician's A-scan display.

Computed radiography is the gold standard at complex joints with extreme geometry, embedded metallic inserts, or thick build-ups, where ultrasound struggles (Spyder Composites). X-ray finds trans-laminar cracks, impact stress fractures, fiber misorientation, wrinkles, and resin-to-fiber ratio inconsistencies. The trade-off is that the equipment is far less portable than an ultrasonic flaw detector.

Active thermography applies a brief thermal pulse to the surface and tracks cooling with an infrared camera. Air pockets, voids, and delaminated plies act as thermal barriers, producing distinct thermal patterns over compromised zones. Thermography scans a wider area per pass than point-contact ultrasound and suits large flat sections like down tubes and top tubes. Presidio Composites runs this method.

The geometry of the scratch decides which subsurface method applies. A scratch on a long flat section of a down tube is a thermography case. A scratch at a complex junction with metallic inserts is a computed-radiography case. A surface-breaking hairline crack at the seatpost insertion zone is a dye-penetrant case. A scratch on a conventional tube where the question is wall-thickness change under the scratch zone is an ultrasound case.

The trap the protocol is built around#

A clean paint surface guarantees nothing about the laminate underneath. The mechanism is well-documented: under compression, adjacent composite plies can separate from one another while the paint and clearcoat above them remain entirely undisturbed. Subsurface fiber fractures and resin voids can sit beneath an unblemished clearcoat. A frame can therefore look pristine and still be structurally compromised.

This is what makes the scratch question two questions and not one. The clearcoat-versus-laminate question filters surface findings into Serviceable or "needs deeper look." The high-stress-zone question decides whether even a clean surface in a junction zone, after a known impact event, is reason for subsurface NDT.

The mechanical reasoning is short. Carbon plies are aligned along specific load paths with significant structural redundancy, so a frame can lose roughly 40 to 50 percent of its local interlaminar shear strength in a specific area and still feel stiff and responsive under moderate riding load (Certify Cycle). The test ride after spotting an unexplained scratch reads the surrounding healthy plies, not the damaged area. It is not confirmation of soundness, which is why the surface triage and the subsurface NDT both exist as separate steps.

What this means for the owner#

A scratch on a carbon bike frame is at minimum a raking-light examination away from a defensible answer. Most scratches that look minor under good light are minor: paint-and-clearcoat damage on a non-junction surface, no surrounding bubbling, abraded rather than cracked, logged as cosmetic and ridden. Some are not: damage that reaches the laminate, sits in a junction zone, or carries surrounding paint bubbling. Those are the cases the subsurface NDT methods exist for, and the geometry of the scratch decides which method applies.

The shortest version of the rule: clean the area, sweep with raking light at a shallow angle, look closely with a loupe or microscope, and read the result against the four high-stress junctions. If the scratch falls in Case 1, the inspection report would log it as Serviceable and the bike rides. If it falls in Case 2, book the subsurface scan before riding again. A clean paint surface is not the answer; it is the question the subsurface methods are built around.